Re: [OT?] write C program with UTF16LE
Adam Bozanich wrote: On Mon, 15 Mar 2004, Matthew Seaman wrote: Errr... no. Not even if you're using the Linux devtools to create a Linux executable. You're using the FreeBSD system libc. Same API, different code, different licencing terms. Yep, one look at /usr/include/stdio.h proves you right. Thanks... I (obviously) didn't realize that. Now I'm very curious. If BSD has it's own C api, did it at one time have it's own compiler? If so, what happened to it? Well, one can track down the original C compiler written by Kernighan & Ritchie. Somewhere around 1981, one can also find references to PCC, the "Portable C Compiler" by Aho & Johnson, which has largely been replaced by GNU's gcc since then. There's been some effort recently to make FreeBSD go using Intel's icc, as well. Does gcc have to know about the different syscall calling conventions? GCC has to know about the local architecture's calling conventions. GCC does not need to know anything special about the system call interface defined by the kernel and libc. -- -Chuck ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: [OT?] write C program with UTF16LE
On Mon, 15 Mar 2004, Matthew Seaman wrote: > On Mon, Mar 15, 2004 at 02:28:17AM -0800, Adam Bozanich wrote: > > On Mon, 15 Mar 2004, Zhang Weiwu wrote: > > > > but I am the kind of newbie don't know if I am using glibc at all. When I > > > just write > > > #include > > > Am i using the stdio.h from glibc? > > > Yes, on FreeBSD you are using GNU's libc > > Errr... no. Not even if you're using the Linux devtools to create a > Linux executable. You're using the FreeBSD system libc. Same API, > different code, different licencing terms. > Yep, one look at /usr/include/stdio.h proves you right. Thanks... I (obviously) didn't realize that. Now I'm very curious. If BSD has it's own C api, did it at one time have it's own compiler? If so, what happened to it? Does gcc have to know about the different syscall calling conventions? I'm trying to look through /usr/src/lib/ right now, but I don't understand most of what is going on. Thanks a lot, -Adam ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: [OT?] write C program with UTF16LE
On Mon, Mar 15, 2004 at 02:28:17AM -0800, Adam Bozanich wrote: > On Mon, 15 Mar 2004, Zhang Weiwu wrote: > > but I am the kind of newbie don't know if I am using glibc at all. When I > > just write > > #include > > Am i using the stdio.h from glibc? > Yes, on FreeBSD you are using GNU's libc Errr... no. Not even if you're using the Linux devtools to create a Linux executable. You're using the FreeBSD system libc. Same API, different code, different licencing terms. But it makes no practical difference to the task at hand. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. 26 The Paddocks Savill Way PGP: http://www.infracaninophile.co.uk/pgpkey Marlow Tel: +44 1628 476614 Bucks., SL7 1TH UK pgp0.pgp Description: PGP signature
Re: [OT?] write C program with UTF16LE
On Mon, 15 Mar 2004, Adam Bozanich wrote: > > Sorry, There is an error here. fgets() reads the number you specify minus one chars, then null terminates the string. these lines: > > char delbuf[CHUNKSIZE]; > char chunks[CHUNKSIZE * MAX_CHUNK_COUNT]; > Should be: char delbuf[CHUNKSIZE + 1]; char chunks[CHUNKSIZE * MAX_CHUNK_COUNT + 1 ]; -Adam ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: [OT?] write C program with UTF16LE
On Mon, 15 Mar 2004, Zhang Weiwu wrote:
> Hello. Although I write some php/perl script, I don't write C program. Now
> I have a very large text file in UTF16LE format, the rule is strings are
> seperated by numbers. Say
>
> 0300 6100 6200 6300 0400 6700 5400 9800 7400 0300
>
> Leading 0300 means the following 3 characters (6 bytes) is a string, and
> the next 0400 means the following 4 characters makes another string.
>
Here's an example using the fgets function. see 'man fgets'. There are
probably a bunch of ways to go about this, but this one is nice and
simple.
#include
#define CHUNKSIZE 5 /* 4 characters and a space */
/* max number of encoded chars if you are using 2 decimal places for the count*/
#define MAX_CHUNK_COUNT 99
int main(int argc, char **argv) {
char delbuf[CHUNKSIZE];
char chunks[CHUNKSIZE * MAX_CHUNK_COUNT];
int chunk_count;
while(fgets(delbuf,CHUNKSIZE+1,stdin) != NULL)
{
/* you may not want to destroy this */
delbuf[2] = '\0';
chunk_count = atoi(delbuf);
if(fgets(chunks, (CHUNKSIZE * chunk_count) + 1 , stdin) == NULL){
fprintf(stderr,"can't read all of the string\n");
break;
}
fprintf(stdout,"\n%s",chunks);
}
exit(0);
}
This worked for the numbers you gave, but I'm sure that you need to add
some better error handling and what not. You probably also don't want
to trash the buffer holding the string length.
Try running with this:
./a.out < inputfile > outputfile
> but I am the kind of newbie don't know if I am using glibc at all. When I
> just write
> #include
> Am i using the stdio.h from glibc?
>
Yes, on FreeBSD you are using GNU's libc
I hope this gives you some ideas, good luck!
-Adam
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
[OT?] write C program with UTF16LE
Hello. Although I write some php/perl script, I don't write C program. Now I have a very large text file in UTF16LE format, the rule is strings are seperated by numbers. Say 0300 6100 6200 6300 0400 6700 5400 9800 7400 0300 Leading 0300 means the following 3 characters (6 bytes) is a string, and the next 0400 means the following 4 characters makes another string. I need to read the file and replace every number-style string seperator with a linefeed. I decide to use C, it is a good chance to start some practice on C. The old getc() I learnt from school is not my cup of tea, because I always need to do two getcs at once, and for the seperators I need to do getc()+getc()*256. What is the best practics to deal with such number/UTF16 mixed text? I googled around and find some tutorials, most i18n toturials think I'm already a C expert:( I find the glibc manual looks good learning resource, but I am the kind of newbie don't know if I am using glibc at all. When I just write #include Am i using the stdio.h from glibc? I think simply point me a tutorial that fits me will do me more help. Thank you. _ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
